# Waniwani

> The AI distribution platform. Build, launch, and optimize your presence across ChatGPT, Claude, Gemini, and every AI platform where customers already ask for help.

Waniwani is the infrastructure layer for the AI economy. We help companies build, deploy, and optimize AI apps, lightweight services built on the MCP (Model Context Protocol) that represent your product inside AI conversations. Think of it as your storefront for the AI economy. We released the first insurance app in ChatGPT.

## Key Platform Features

### AI App Builder
Build and deploy your AI app from product knowledge ingestion to multi-platform distribution. No AI expertise or technical team required. Can be live in two weeks. Your AI app answers customer questions, generates quotes, and captures leads directly inside AI conversations.

### Synthetic Buyer Monitoring
Deploy synthetic buyer agents that simulate real customer journeys across all major AI platforms. Test whether AI recommends your product, identify which competitors appear instead, and get actionable insights to improve your visibility. Supports A/B testing with different buyer personas.

### AI App Analytics
Full-funnel visibility into how your AI app performs across every platform. Track call volume, conversion rates, lead capture, and drop-off points. See which queries drive revenue, all consolidated in one real-time dashboard.

### Compliance Monitoring
Continuous scanning of your AI distribution for regulatory compliance. Covers GDPR, EU AI Act, DORA, SOC 2, and other evolving regulations. Infrastructure automatically adapts as standards change, so you don't have to scramble.

## Solutions by Audience

### For Insurance & Financial Services
Distribute your products across AI platforms where customers already ask for coverage and financial advice. Generate quotes, capture leads, and stay compliant with industry-specific regulations including DORA and Solvency II.

### For Marketing & Growth Teams
Turn AI platforms into a high-performing acquisition channel. Monitor how AI models recommend your product, optimize your visibility with synthetic buyer testing, and measure conversion from AI conversations.

### For Product & Engineering Teams
Build your own AI app on our infrastructure layer. We handle distribution across platforms, analytics, synthetic buyer testing, and compliance monitoring. You focus on building, we handle the rest.

## Technical Capabilities and Integrations

### AI Platforms Supported
ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity, and custom website integrations. New platforms are added automatically as they emerge.

### Architecture
- Top layer: All major AI platforms plus your own website
- Middle layer: Your AI app (quotes, leads, answers)
- Foundation layer: Waniwani infrastructure (build, optimize, measure, distribute)

### Protocol
Built on MCP (Model Context Protocol) for native integration with AI platforms.

## Compliance & Security

- SOC 2 Type II Certified
- GDPR Compliant
- EU AI Act Ready
- DORA Compliant

## Industries

Waniwani serves companies in regulated and high-consideration industries, including insurance, financial services, fintech, and healthcare.

## FAQ

### What's an AI app?
A lightweight service built on the MCP protocol, designed to represent your product inside AI conversations. It answers questions, generates quotes, and captures leads.

### What if I don't have an AI app yet?
We build it for you, from product knowledge ingestion to multi-platform distribution. No AI expertise needed.

### What if I want to build my own AI app?
We provide the infrastructure layer: distribution across platforms, analytics, synthetic buyer testing, and compliance monitoring.

### How do you monitor what AI says about my product?
Synthetic buyers simulate real customer journeys across every AI platform, tracking whether AI recommends you and providing actionable insights to improve visibility.

### How do you handle compliance?
Continuous monitoring for GDPR, EU AI Act, DORA, and other evolving regulations. When standards shift, we adapt your infrastructure automatically.

## Open Source

@waniwani/sdk is our open-source (MIT) TypeScript SDK for building distribution MCPs: multi-step conversational funnels (lead generation, booking, quotes) that run as a single MCP tool inside ChatGPT, Claude, and Cursor. A typed state graph compiles to one MCP tool. Bring your own state backend; an optional hosted tier (event tracking, funnel analytics, knowledge base, chat widget) is opt-in via a single env var. Full agent-readable OSS docs live at https://docs.waniwani.ai/llms.txt.

- SDK: https://www.npmjs.com/package/@waniwani/sdk
- Repo: https://github.com/WaniWani-AI/sdk
- Template: https://github.com/WaniWani-AI/mcp-distribution-template
- Docs (OSS llms.txt): https://docs.waniwani.ai/llms.txt
- Coding-agent skill: npx skills add WaniWani-AI/sdk -s waniwani-sdk

## Developer Resources

Everything below is public, needs no login, and lives at a URL that does not move. The
human-readable index is at https://waniwani.ai/developers and the markdown twin at
https://waniwani.ai/developers.md.

### Waniwani API documentation

- [Waniwani developer documentation](https://docs.waniwani.ai): SDK guides, MCP server reference, platform configuration, deployment.
- [Waniwani docs llms.txt](https://docs.waniwani.ai/llms.txt): Agent-readable index of every documentation page.
- [Waniwani docs llms-full.txt](https://docs.waniwani.ai/llms-full.txt): The whole documentation set as one markdown file.
- [Waniwani SDK quickstart](https://docs.waniwani.ai/sdk/quickstart): A flow running in under five minutes.

### Waniwani OpenAPI specification

- [Waniwani website API OpenAPI spec](https://waniwani.ai/openapi.json): OpenAPI 3.1 for the public endpoints on waniwani.ai, including the shared JSON error model.
- [Waniwani platform API reference](https://docs.waniwani.ai/api-reference/openapi.json): OpenAPI document behind the developer portal's API reference.
- [Waniwani website API index](https://waniwani.ai/api): JSON service index of every endpoint on this domain.

### Waniwani MCP server

Hosted Model Context Protocol server over Streamable HTTP, OAuth 2.1 with PKCE, four
tools: whoami, set_active_org, search, execute.

- [Waniwani MCP server (US)](https://mcp.waniwani.ai/mcp): US data residency.
- [Waniwani MCP server (EU)](https://eu.mcp.waniwani.ai/mcp): EU data residency.
- [Waniwani MCP server card](https://waniwani.ai/.well-known/mcp/server-card.json): Discovery manifest with endpoints, transport, auth, and the tool list.
- [Waniwani MCP server documentation](https://docs.waniwani.ai/mcp-server/overview): What the server does and example prompts.
- [Connect an AI client](https://docs.waniwani.ai/mcp-server/connect): Setup for claude.ai, Claude Code, Cursor, and ChatGPT.

### Waniwani authentication

- [Waniwani OAuth protected resource metadata](https://mcp.waniwani.ai/.well-known/oauth-protected-resource): RFC 9728 document naming the authorization server and supported scopes.
- [Waniwani MCP scopes and consent](https://docs.waniwani.ai/mcp-server/how-it-works): How scopes are checked and what the sandboxed execute tool can reach.
- [Waniwani API key setup](https://docs.waniwani.ai/sdk/configuration/api-key): Create a WANIWANI_API_KEY on the free tier.
- [Waniwani security and key management](https://docs.waniwani.ai/sdk/configuration/security): Encryption at rest, key management, rotation.

### Waniwani SDK and CLI

- [@waniwani/sdk on npm](https://www.npmjs.com/package/@waniwani/sdk): TypeScript SDK for building MCP funnels.
- [@waniwani/cli on npm](https://www.npmjs.com/package/@waniwani/cli): Wire a local repo to a Waniwani agent.
- [Waniwani SDK on GitHub](https://github.com/WaniWani-AI/sdk): Source, issues, and the AGENTS.md rules for coding agents.
- [Waniwani MCP distribution template](https://github.com/WaniWani-AI/mcp-distribution-template): Starter repo for a production distribution MCP.
- [Waniwani agent skills](https://skills.sh/waniwani-ai): Published skills for coding agents.

### Error handling

Every endpoint on waniwani.ai fails as JSON, never as an HTML error page:

    {
      "error": {
        "code": "missing_field",
        "message": "The \"email\" field is required.",
        "hint": "Include a work email address, e.g. {\"email\":\"you@company.com\"}.",
        "status": 422,
        "documentation_url": "https://waniwani.ai/developers#api-documentation",
        "field": "email"
      },
      "request_id": "req_1a4b7c9e2d5f8036b1c4e7a9"
    }

Branch on `error.code`. The codes are bad_request, invalid_json, invalid_email,
missing_field, not_found, method_not_allowed, unsupported_media_type, internal_error.

## Links

- [Homepage](https://www.waniwani.ai)
- [Developers](https://waniwani.ai/developers)
- [Blog](https://www.waniwani.ai/blog)
- [Request a Demo](https://www.waniwani.ai)